Current strategies aiming at identifying protective cytotoxic T cell (CTL) responses and designing vaccines against HIV are based on the assumption that immune responses elicited during HIV infection are only directed against HIV epitopes. The limitations of these strategies are the elusiveness of HIV-specific protective immune responses and the fact that many HIV-specific CTL exert pressure on HIV leading to viral mutations and subsequent immune escape. Thus eliciting HIV-specific CTL responses may not be sufficient to block the transmission of HIV. This proposal will specifically test the hypothesis that unconventional HIV-induced host- derived neoepitopes uniquely processed and presented in HIV-infected cells (but not in healthy cells) elicit CTL responses contributing to spontaneous HIV immune control. The identification of these HIV-induced Host- Derived Antiviral Epitopes (HDAE) and corresponding CTL responses -not subjected to immune pressure- will be performed in a unique cohort of HIV spontaneous controllers with weak HIV-specific CTL responses. Cellular and HIV proteins are degraded into epitopes or epitope precursors by the proteasome and other peptidases in the cytosol and then trimmed in the endoplasmic reticulum before being loaded onto MHC-I and displayed at the cell surface for recognition by CTL. Building on novel epitope processing assays, we showed preferential processing of some HIV epitopes, a property that relies on motifs we used to alter the production of irrelevant epitopes. We also identified a novel factor involved in epitope processing efficiency, namely the highly variable intracellular stability of optimal HIV epitopes, also driven by specific motifs. In HIV-infected cells, the altered expression and activities of the antigen processing machinery that we recently identified, the presence of an additional HIV-encoded protease along with massive degradation of specific cellular targets are likely to alter the degradation pattern of cellular proteins and produce HDAE. Yet it is impossible to identify HDAE-specific CTL responses by a comprehensive screen of cellular peptides. Through a collaboration with Microsoft Research, we will build a customized prediction program and scan the human genome for antigenic peptides produced in HIV-infected primary cells. With an innovative mass spectrometry- based analysis, we will identify antigenic precursors uniquely found in the cytosol of HIV-infected CD4 T cells from HIV controllers. Their identity will be confirmed through computational analysis and in vitro degradation of target proteins. We will screen HIV-infected persons for CTL responses against HDAE, isolate HDAE-specific CTL and assess their antiviral capacity. We will also assess the capacity of monocyte-derived dendritic cells to cross-present HDAE endogenously processed by HIV-infected CD4 T cells, to stimulate CTL whose antiviral capacity will be assessed. This proposal relies on a cross-disciplinary collaboration involving computational science, novel biochemical and immunological assays designed for primary cells. Identifying HDAE will contribute to the design of novel immunogens eliciting CTL responses that will prevent HIV transmission.